Soybean oil accounts for about 70% of the 14 billion pounds of edible oil consumed in the United States and is a major edible oil worldwide. It is used in baking, frying, salad dressing, margarine, and a multitude of processed foods. In 1987/88 60 million acres of soybean were planted in the U.S. Soybean is the lowest-cost producer of vegetable oil, which is a by-product of soybean meal. Soybean is agronomically well-adapted to many parts of the U.S. Machinery and facilities for harvesting, storing, and crushing are widely available across the U.S. Soybean products are also a major element of foreign trade since 30 million metric tons of soybeans, 25 million metric tons of soybean meal, and 1 billion pounds of soybean oil were exported in 1987/88. Nevertheless, increased foreign competition has lead to recent declines in soybean acreage and production. The low cost and ready availability of soybean oil provides an excellent opportunity to upgrade this commodity oil into higher value speciality oils to both add value to soybean crop for the U.S. farmer and enhance U.S. trade.
Soybean oil derived from commercial varieties is composed primarily of 11% palmitic (16:0), 4% stearic (18:0), 24% oleic (18:1), 54% linoleic (18:2) and 7% linolenic (18:3) acids. Palmitic and stearic acids are, respectively, 16- and 18-carbon-long saturated fatty acids. Oleic, linoleic and linolenic are 18-carbon-long unsaturated fatty acids containing one, two and three double bonds, respectively. Oleic acid is also referred to as a monounsaturated fatty acid, while linoleic and linolenic acids are also referred to as polyunsaturated fatty acids. The specific performance and health attributes of edible oils is determined largely by their fatty acid composition.
Soybean oil is high in saturated fatty acids when compared to other sources of vegetable oil and contains a low proportion of oleic acid, relative to the total fatty acid content of the soybean seed. These characteristics do not meet important health needs as defined by the American Heart Association.
More recent research efforts have examined the role that monounsaturated fatty acid plays in reducing the risk of coronary heart disease. In the past, it was believed that monounsaturates, in contrast to saturates and polyunsaturates, had no effect on serum cholesterol and coronary heart disease risk. Several recent human clinical studies suggest that diets high in monounsaturated fat may reduce the “bad” (low-density lipoprotein) cholesterol while maintaining the “good” (high-density lipoprotein) cholesterol. [See Mattson et al. (1985) Journal of Lipid Research 26:194-202, Grundy (1986) New England Journal of Medicine 314:745-748, and Mensink et al. (1987) The Lancet 1:122-125, all collectively herein incorporated by reference.] These results corroborate previous epidemiological studies of people living in Mediterranean countries where a relatively high intake of monounsaturated fat and low consumption of saturated fat correspond with low coronary heart disease mortality. [Keys, A., Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease, Cambridge: Harvard University Press, 1980, herein incorporated by reference.] The significance of monounsaturated fat in the diet was further confirmed by international researchers from seven countries at the Second Colloquim on Monounsaturated Fats held Feb. 26, 1987, in Bethesda, Md., and sponsored by the National Heart, Lung and Blood Institutes [Report, Monounsaturates Use Said to Lower Several Major Risk Factors, Food Chemical News, Mar. 2, 1987, p. 44, herein incorporated by reference].
Soybean oil is also relatively high in polyunsaturated fatty acids—at levels in far excess of our essential dietary requirement. These fatty acids oxidize readily to give off-flavors and result in reduced performance associated with unprocessed soybean oil. The stability and flavor of soybean oil is improved by hydrogenation, which chemically reduces the double bonds. However, the need for this processing reduces the economic attractiveness of soybean oil.
A soybean oil low in total saturates and polyunsaturates and high in monounsaturate would provide significant health benefits to the United States population, as well as, economic benefit to oil processors. Soybean varieties which produce seeds containing the improved oil will also produce valuable meal as animal feed.
Another type of differentiated soybean oil is an edible fat for confectionary uses. More than 2 billion pounds of cocoa butter, the most expensive edible oil, are produced worldwide. The U.S. imports several hundred million dollars worth of cocoa butter annually. The high and volatile prices and uncertain supply of cocoa butter have encouraged the development of cocoa butter substitutes. The fatty acid composition of cocoa butter is 26% palmitic, 34% stearic, 35% oleic and 3% linoleic acids. About 72% of cocoa butter's triglycerides have the structure in which saturated fatty acids occupy positions 1 and 3 and oleic acid occupies position 2. Cocoa butter's unique fatty acid composition and distribution on the triglyceride molecule confer on it properties eminently suitable for confectionary end-uses: it is brittle below 27° C. and depending on its crystalline state, melts sharply at 25-30° C. or 35-36° C. Consequently, it is hard and non-greasy at ordinary temperatures and melts very sharply in the mouth. It is also extremely resistant to rancidity. For these reasons, producing soybean oil with increased levels of stearic acid, especially in soybean lines containing higher-than-normal levels of palmitic acid, and reduced levels of unsaturated fatty acids is expected to produce a cocoa butter substitute in soybean. This will add value to oil and food processors as well as reduce the foreign import of certain tropical oils.
Only recently have serious efforts been made to improve the quality of soybean oil through plant breeding, especially mutagenesis, and a wide range of fatty acid composition has been discovered in experimental lines of soybean (Table 1). These findings (as well as those with other oilcrops) suggest that the fatty acid composition of soybean oil can be significantly modified without affecting the agronomic performance of a soybean plant. However, there is no soybean mutant line with levels of saturates less than those present in commercial canola, the major competitor to soybean oil as a “healthy” oil.
TABLE 1Range of Fatty Acid PercentagesProduced by Soybean MutantsRange ofFatty AcidsPercentagesPalmitic Acid6-28Stearic Acid3-30Oleic Acid17-50 Linoleic Acid35-60 Linolenic Acid3-12
There are serious limitations to using mutagenesis to alter fatty acid composition. It is unlikely to discover mutations a) that result in a dominant (“gain-of-function”) phenotype, b) in genes that are essential for plant growth, and c) in an enzyme that is not rate-limiting and that is encoded by more than one gene. Even when some of the desired mutations are available in soybean mutant lines their introgression into elite lines by traditional breeding techniques will be slow and expensive, since the desired oil compositions in soybean are most likely to involve several recessive genes.
Recent molecular and cellular biology techniques offer the potential for overcoming some of the limitations of the mutagenesis approach, including the need for extensive breeding. Particularly useful technologies are: a) seed-specific expression of foreign genes in transgenic plants [see Goldberg et al.(1989) Cell 56:149-160], b) use of antisense RNA to inhibit plant target genes in a dominant and tissue-specific manner [see van der Krol et al. (1988) Gene 72:45-50], c) transfer of foreign genes into elite commercial varieties of commercial oilcrops, such as soybean [Chee et al. (1989) Plant Physiol. 91:1212-1218; Christou et al. (1989) Proc. Natl. Acad. Sci. U.S.A. 86:7500-7504; Hinchee et al. (1988) Bio/Technology 6:915-922; EPO publication 0 301 749 A2], rapeseed [De Block et al. (1989) Plant Physiol. 91:694-701], and sunflower [Everett et al.(1987) Bio/Technology 5:1201-1204], and d) use of genes as restriction fragment length polymorphism (RFLP) markers in a breeding program, which makes introgression of recessive traits into elite lines rapid and less expensive [Tanksley et al. (1989) Bio/Technology 7:257-264]. However, application of each of these technologies requires identification and isolation of commercially-important genes.
Oil biosynthesis in plants has been fairly well-studied [see Harwood (1989) in Critical Reviews in Plant Sciences, Vol. 8(1):1-43]. The biosynthesis of palmitic, stearic and oleic acids occur in the plastids by the interplay of three key enzymes of the “ACP track”: palmitoyl-ACP elongase, stearoyl-ACP desaturase and acyl-ACP thioesterase. Stearoyl-ACP desaturase introduces the first double bond on stearoyl-ACP to form oleoyl-ACP. It is pivotal in determining the degree of unsaturation in vegetable oils. Because of its key position in fatty acid biosynthesis it is expected to be an important regulatory step. While the enzyme's natural substrate is stearoyl-ACP, it has been shown that it can, like its counterpart in yeast and mammalian cells, desaturate stearoyl-CoA, albeit poorly [McKeon et al. (1982) J. Biol. Chem. 257:12141-12147]. The fatty acids synthesized in the plastid are exported as acyl-CoA to the cytoplasm. At least three different glycerol acylating enzymes (glycerol-3-P acyltransferase, 1-acyl-glycerol-3-P acyltransferase and diacylglycerol acyltransferase) incorporate the acyl moieties from the cytoplasm into triglycerides during oil biosynthesis. These acyltransferases show a strong, but not absolute, preference for incorporating saturated fatty acids at positions 1 and 3 and monounsaturated fatty acid at position 2 of the triglyceride. Thus, altering the fatty acid composition of the acyl pool will drive by mass action a corresponding change in the fatty acid composition of the oil. Furthermore, there is experimental evidence that, because of this specificity, given the correct composition of fatty acids, plants can produce cocoa butter substitutes [Bafor et al. (1990) JAOCS 67:217-225].
Based on the above discussion, one approach to altering the levels of stearic and oleic acids in vegetable oils is by altering their levels in the cytoplasmic acyl-CoA pool used for oil biosynthesis. There are two ways of doing this genetically: a) altering the biosynthesis of stearic and oleic acids in the plastid by modulating the levels of stearoyl-ACP desaturase in seeds through either overexpression or antisense inhibition of its gene, and b) converting stearoyl-CoA to oleoyl-CoA in the cytoplasm through the expression of the stearoyl-ACP desaturase in the cytoplasm.
In order to use antisense inhibition of stearoyl-ACP desaturase in the seed, it is essential to isolate the gene(s) or cDNA(s) encoding the target enzyme(s) in the seed, since antisense inhibition requires a high-degree of complementarity between the antisense RNA and the target gene that is expected to be absent in stearoyl-ACP desaturase genes from other species.
The purification and nucleotide sequences of mammalian microsomal stearoyl-CoA desaturases have been published [Thiede et al. (1986) J. Biol. Chem. 262:13230-13235; Ntambi et al. (1988) J. Biol. Chem. 263:17291-17300; Kaestner et al. (1989) J. Biol. Chem. 264:14755-14761]. However, the plant enzyme differs from them in being soluble, in utilizing a different electron donor, and in its substrate-specificities. The purification and the nucleotide sequences for animal enzymes do not teach how to purify the plant enzyme or isolate a plant gene. The purification of stearoyl-ACP desaturase was reported from safflower seeds [McKeon et al. (1982) J. Biol. Chem. 257:12141-12147]. However, this purification scheme was not useful for soybean, either because the desaturases are different or because of the presence of other proteins such as the soybean seed storage proteins in seed extracts.
The rat liver stearoyl-CoA desaturase protein has been expressed in E. coli [Strittmatter et al. (1988) J. Biol. Chem. 263:2532-2535] but, as mentioned above, its substrate specificity and electron donors are quite distinct from that of the plant.
Plant stearoyl-ACP desaturase cDNAs have been cloned from soybean [U.S. Pat. No. 5,760,206, the disclosure of which is hereby incorporated by reference], safflower [Thompson et al. (1991) Proc. Natl. Acad. Sci. 88:2578], castor [Shanklin and Somerville (1991) Proc. Natl. Acad. Sci. 88:2510-2514], and cucumber [Shanklin et al. (1991) Plant Physiol. 97:467-468]. Kutzon et al. [(1992) Proc. Natl. Acad. Sci. 89:2624-2648] have reported that rapeseed stearoyl-ACP desaturase when expressed in Brassica rapa and B. napa in an antisense orientation can result in increase in 18:0 level in transgenic seeds. All of the reported genes have 59-80% identity to each other at the nucleotide and polypeptide level.
U.S. Pat. No. 5,723,595, issued to Thompson et al. on Mar. 3, 1998, describes stearoyl-ACP desaturases from castor and safflower.
U.S. Pat. No. 5,443,974, issued to Hitz et al., on Aug. 22, 1995, describes soybean stearoyl-ACP desaturase.
U.S. Pat. No. 5,760,206, issued to Hitz et al, on Jun. 2, 1998, describes soybean stearoyl-ACP desaturase.